Liu J C, Shiffman R N
Center for Medical Informatics, Yale University School of Medicine, New Haven, Connecticut, USA.
Proc AMIA Annu Fall Symp. 1997:283-7.
There are a number of obstacles to successful operationalization of clinical practice guidelines, including the difficulty in accurately representing a statement's decidability or an action's executability. Both require reasoning with incomplete and imprecise information, and we present one means of processing such information. We begin with a brief overview of fuzzy set theory, in which elements can have partial memberships in multiple sets. With fuzzy inferencing, these sets can be combined to create multiple conclusions, each with varying degrees of truth. We demonstrate a fuzzy model developed from a published clinical practice guideline on the management of first simple febrile seizures. Although the creation of fuzzy sets can be an arbitrary process, we believe that fuzzy inferencing is an effective tool for the expression of guideline recommendations, and that it can be useful for the management of imprecision and uncertainty.
临床实践指南的成功实施存在诸多障碍，包括难以准确体现陈述的可判定性或行动的可执行性。这两者都需要运用不完整和不精确的信息进行推理，我们提出了一种处理此类信息的方法。我们首先简要概述模糊集理论，在该理论中，元素可以部分属于多个集合。通过模糊推理，这些集合可以组合起来得出多个结论，每个结论都有不同程度的真实性。我们展示了一个基于已发表的关于首次单纯性热性惊厥管理的临床实践指南开发的模糊模型。虽然模糊集的创建可能是一个随意的过程，但我们认为模糊推理是表达指南建议的有效工具，并且它对于处理不精确性和不确定性可能是有用的。
